Spaces:
Sleeping
Sleeping
File size: 2,890 Bytes
b918d4f e49f323 b918d4f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
import streamlit as st
import os
from langchain import HuggingFaceHub, PromptTemplate, LLMChain
# Define your AI assistant setup and configurations here
os.environ['API_KEY'] = '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'
model_id = 'tiiuae/falcon-7b-instruct'
falcon_llm = HuggingFaceHub(huggingfacehub_api_token=os.environ['API_KEY'],
repo_id=model_id,
model_kwargs={"temperature": 0.8, "max_new_tokens": 2000})
template = """
You are an AI assistant that provides helpful answers to user queries.
{conversation}
"""
prompt = PromptTemplate(template=template, input_variables=['conversation'])
falcon_chain = LLMChain(llm=falcon_llm,
prompt=prompt,
verbose=True)
# Define the Streamlit app
def main():
st.title("Mouli's AI Assistant")
# Initialize conversation history as a list
conversation_history = st.session_state.get("conversation_history", [])
# Create an input box at the bottom for user's message
user_message = st.text_input("Your message:")
# If the user's message is not empty, process it
if user_message:
# Add user's message to conversation history
conversation_history.append(("user", user_message))
# Combine conversation history to use as input for the AI assistant
conversation_input = "\n".join([f"{author}: {message}" for author, message in conversation_history])
# Use your AI assistant to generate a response based on the conversation
response = falcon_chain.run(conversation_input)
# Add AI's response to conversation history
conversation_history.append(("AI", response))
# Store the updated conversation history in session state
st.session_state.conversation_history = conversation_history
# Display the conversation history
display_conversation(conversation_history)
def display_conversation(conversation_history):
st.markdown("<style>.message-container { display: flex; flex-direction: row; padding: 16px; }</style>", unsafe_allow_html=True)
st.markdown("<style>.user-message { align: left; background-color: green; padding: 8px; border-radius: 8px; margin: 4px; }</style>", unsafe_allow_html=True)
st.markdown("<style>.ai-message { align: right; background-color: black; padding: 8px; border-radius: 8px; margin: 4px; }</style>", unsafe_allow_html=True)
st.markdown("<div class='message-container'>", unsafe_allow_html=True)
for author, message in conversation_history:
if author == "AI":
st.markdown(f"<div class='ai-message'>{message}</div>", unsafe_allow_html=True)
else:
st.markdown(f"<div class='user-message'>{message}</div>", unsafe_allow_html=True)
st.markdown("</div>", unsafe_allow_html=True)
if __name__ == "__main__":
main()
|